worthwhile to repeat their experiment, but with fuller documentation. It is wrong to
assume that all we know about continental vertebrates in the Late Cretaceous comes from
North America. Sea levels were high worldwide, but continental units with fossil
vertebrates are known (Weishampel 1990) from around the Mediterranean (Portugal,
Spain, France, Romania, Middle East), from Asia (Uzbekistan, Tadzhikistan, Kazakhstan,
Russia, India, Mongolia, China, Japan, Laos), from South America (Colombia, Bolivia,
Peru, Brazil, Uruguay, Chile, Argentina), and from Africa (Morocco, Algeria, Egypt,
Niger, Kenya, South Africa, Madagascar). Mammal fossils are known from all these areas,
with spectacular examples from numerous horizons in the Late Cretaceous of Uzbekistan,
Kazakhstan, Mongolia, China, Argentina, and Madagascar. Expanding the sampling
worldwide might not in fact invalidate the findings by Foote et al. (1999).
Oldest fossilsThe current literature about origins of major clades often includes discussion about the
oldest relevant fossils. Does close scrutiny of the fossil evidence indicate rapprochement
between molecular early dates and palaeontological late dates of origin? This does not
seem to be the case for the earliest metazoans, green plants, angiosperms, or birds, but
there is now good agreement for the basal radiation of mammals.
MismatchThe first metazoan fossils are generally accepted to date from the earliest Cambrian, the
great burst of expansion of skeletonized animal groups from about 545 Ma, long after the
molecular date of around 1000 Ma. Fossil evidence for metazoans is known from the
Precambrian, but nothing before about 600 Ma. This evidence includes possible sponges
and cnidarians from the Ediacara faunas, exquisitely preserved fossil embryos, and simple
creeping trails (Valentine et al. 1999; Conway Morris 2000).
The first vascular land plants are found as fossils in the Silurian, and earlier evidence
from spores extends the range back to the Ordovician (475 Ma; Kenrick and Crane
1997), considerably younger than a molecular estimate of 700 Ma (Heckman et al. 2001).
A similar gap exists for angiosperms, with the oldest generally accepted fossils being from
the Early Cretaceous, pollen records dated at about 130 Ma, and abundant pollen and
macroplant fossils from 120 Ma onwards (Crane et al. 1995). Older putative fossil
angiosperms, from the Jurassic, and even from the Triassic, have not been generally
accepted. DNA sequence evidence places the divergence of angiosperms in the Mid-Jurassic
(175 Ma; Wikström et al. 2001). However, genealogical evidence actually suggests a
much more ancient date of origin, back in the Carboniferous at 290 Ma (Kenrick 1999), if
it turns out that the sister group of angiosperms is the gymnosperms.
In the case of the origin of modern birds, many supposed Cretaceous representatives of
modern bird orders have been cited (e.g. Cooper and Penny 1997; van Tuinen et al.
2000; Paton et al. 2002), but most have been disputed, because the fossils are generally
isolated elements (Dyke 2001; Chiappe and Dyke 2002). Hence, the oldest
uncontroversial fossils of modern bird orders date from the latest Cretaceous or
MICHAEL J.BENTON 75